digest
Short Summary
Main recommendation: stay invested through market crashes and keep to the asset allocation you set during calmer times, rather than selling because a scary narrative makes the decline feel different. Ben Felix argues that historical data show sharp declines are typically followed by large positive returns, and that selling risks missing those rebounds. He also notes that narratives can distort how investors interpret the same facts.
Key Points
- Stock prices reflect expectations about future profits and risk, so sharp declines are expected from time to time and should be built into every investment plan.
- Narratives can change how people respond to facts through the framing effect and representativeness bias, even though they do not change the facts themselves, as Robert Shiller describes in "Narrative Economics" (2017).
- Goetzmann and Kim's 2017 study of 1,032 crashes of more than 50% over 12 months across 101 global markets (1692 to 2015) found a higher probability of large positive returns after a crash.
- Reducing equity allocation after a crash may be one of the worst moves, because the biggest return months often occur during or after large drawdowns.
- Over the 2000 to 2020 period, which included the dotcom crash, 9/11, SARS, the global financial crisis, and COVID-19, the MSCI All Country World Index still averaged 6.56% annually in Canadian dollar terms.
Details Worth Noting
- Concrete instructions:
- Before a downturn, set an asset allocation you can stick with in both good times and bad, and keep to it during declines.
- Keep an emergency fund so you do not need to sell stocks to cover expenses during a downturn.
- If a drop has been too painful to live through, ask whether you were taking too much risk to begin with, and consider adjusting the allocation during calmer periods rather than selling in a crash.
- Avoid getting out of the market to dodge a drop, because getting back in at the right time is difficult and missing the best months sharply lowers returns.
- For the S&P 500 from 1926 through February 2020, the compound annual return was 10.08%, but removing only the top 10 months (about 0.9% of months) dropped it to 7.74%.
- Historical example: after the 1920 to 1921 recession, the S&P 500 cyclically adjusted P/E hit its lowest level on record in December 1920, then prices rebounded, and they rebounded again after the 1929 decline.
- Shiller defines a bubble as a social epidemic involving extravagant expectations about the future; he won the 2013 Nobel Prize in economics for his work on asset price bubbles.
- Cost estimates cited in the video: the World Bank estimated a 1918-style pandemic would cost about $3 trillion USD, and the IMF estimated the 2007 to 2008 financial crisis write-downs at about $4 trillion USD.
Bottom Line
Market crashes are normal and historically followed by recoveries, so the discipline that matters most is holding a plan built for risk you can actually tolerate, not predicting when the narrative will turn.
